Upgrade is broken again :-(

Perhaps blaming the language for the programmer's mistake makes no sense.

You say it's a backtrace.  In Python that happens when someone didn't
anticipate some exception or chose not to catch it.

If you're going to suggest that the choice of programming language is somehow wrong then
you'll have to explain just exactly what this backtrace is and why a competent programmer
would have been unable to avoid it.  (For example if the backtrace pointed to a bug in one of
the Python standard libraries, or if some bug in the runtime/interpreter triggered it, or
triggered some exceptional condition such as OOM, that, in turn, cause this.

Are you seriously suggesting that higher quality of code would have been likely if they had
chosen C, C++, Perl?

(Personally I have found that the code in Anaconda is just about the worst quality Python
programming I've never seen.  Their kickstart parser is fragile and completely inadequate even
for the limited exposure to sysadmins.  But I don't blame that on their choice of language.
Also the whole process is so tightly coupled together that it seems impossible for them to
support something as simple as: %pre does all my partitioning, filesystem formatting, and
mounting --- everything is on /mnt/sysimage and I've generated my own fstab; skip to the
package installation!)

(If they had than then there'd be no need for an rpmstrap command --- you'd just run anaconda
with the correct "--skip" argument and your virtual iSCSI installation would be all done where
you're Xen, KVM, VMWare or whatever could go pick it up and use it).


Bah.

--
JimD